3. PPM & DS, Hydroacoustic Monitoring Report, USACE MCR Sand Island Test Piles, RMDT, Jan 18 2021 (RMDT 2021).pdf

Robert Miner Dynamic Testing, Inc.

Dynamic Measurements and Analyses for Deep Foundations

Hydroacoustic Monitoring Report

Sand Island Test Pile Installation and Removal Clatsop County, Oregon

September 16- October 15, 2020

Prepared for Pacific Pile and Marine, LLC and Duwamish Services, LLC

For submission to US Army Corps of Engineers, Portland District

Prepared by Robert Miner Dynamic Testing, Inc.

January, 2020

Mailing Address: P.O. Box 340, Manchester, WA, 98353, USA Phone: 360-871-5480 Location: 2288 Colchester Dr. E., Ste A, Manchester, WA, 98353 Fax: 360-871-5483

Table of Contents

1.0 Introduction

1.1 Project Summary and Setting

1.2 Piles

1.3 Hammers and Driving Operations

2.0 Methods

2.1 Monitoring Locations

2.2 Monitoring Equipment

2.3 Measurement Sequence

2.4 Analyses

2.4.1 Analyses for Vibratory Driving

2.4.2 Analyses for Impact Driving

2.4.3 Spectral Considerations

2.4.4 Transmission Loss

3.0 Results

3.1 Characteristic Waveforms and Spectra

3.2 Sound Levels for Vibratory Driving

3.2.1 Broadband Near Source Levels, Vibratory Driving

3.2.2 M-Weighted Near Source Levels, Vibratory Driving

3.3 Sound Levels for Impact Driving

3.3.1 Broadband Near Source Levels, Impact Driving

3.3.2 M-Weighted Near Source Levels, Impact Driving

3.4 Transmission Loss

3.5 Background Sound Measurements

4.0 Discussion and Summary

4.1 Near Source Sound Levels

4.1.1 Vibratory Pile Driving

4.1.2 Impact Pile Driving

4.2 Transmission Loss

4.3 Computed Distances to Sound Thresholds

4.3.1 Vibratory Pile Driving

4.3.2 Impact Pile Driving

4.3.3 Extension of Results

4.4 Background Sound

List of Tables:

Table 1a Pile Dike 6.37 locations and Test Details Table 1b Pile Dike 4.01 locations and Test Details Table 2 Pile Driving Hammer and Driving System Details Table 3a VIBRATORY DRIVING, Statistics for Broadband rms SPL, 10 second integration period, and Cumulative SEL for a 10 m radial distance from the pile, Pile Dike 6.37

Table 3b VIBRATORY DRIVING, Statistics for Broadband rms SPL, 10 second integration period, and Cumulative SEL for a 10 m radial distance from the pile, Pile Dike 4.01

Table 4a IMPACT DRIVING, Statistics Broadband Single Strike peak SPL, dB re:1µPa, for a 10 m radial distance from the pile, Pile Dike 6.37. 21

Table 4b IMPACT DRIVING, Statistics Broadband Single Strike peak SPL, dB re:1µPa, for a 10 m radial distance from the pile Pile Dike 4.01 . 22

Table 5a IMPACT DRIVING, Statistics Broadband Single Strike RMS90, dB re:1µPa, for a 10 m radial distance from the pile, Pile Dike 6.37

Table 5b IMPACT DRIVING, Statistics Broadband Single Strike RMS90, dB re:1µPa, for a 10 m radial distance from the pile, Pile Dike 4.01

Table 6a IMPACT DRIVING, Statistics Broadband Single Strike SEL and Daily cSEL, for a 10 m radial distance from the pile, Pile Dike 6.37

Table 6b IMPACT DRIVING, Statistics Broadband Single Strike SEL and Daily cSEL, for a 10 m radial distance from the pile Pile Dike 4.01

Table 7 Transmission Loss Coefficients computed for each Day Table 8 Computed Distances to Marine Mammal Injury Thresholds Table 9 Computed Distances to Marine Mammal Disturbance Thresholds . . 29 Table 10 Computed Distances to Fish Injury Thresholds

List of Figures:

F1 Characteristic Vibratory Time Series Event F2 Characteristic Impact Time Series Event F3 Power Spectral Density for Characteristic Vibratory Event F4 Power Spectral Density for Characteristic Impact F5 Power Spectral Density with M-weighting, Vibratory Event F6 Power Spectral Density with M-weighting, Impact Event F7 Characteristic Time Series and Power Spectral Density, Impact Event . . . 35 F8 Comparison of Measurements at two depths, 100 m F9 Comparison of Measurements at two depths, 500 m F10 Background Sound Measurements for a 3-Day period F11 Cumulative Distribution for Background Sound F12 Background Sound, PSD and One-Third Octave Band F13 Variation of Time Series and PSD results with distance from Source

Appendix A Characteristic Waveforms and Power Spectral Density A1 Vibratory Driving A2 Impact Driving

Appendix B Graphic Summary of Computed Sound Descriptors B1 Vibratory Driving Broadband B2 Impact Driving Broadband

Appendix C Sound Descriptors for Vibratory Driving Broadband and M-weighted, RMS and CSEL

Appendix D Sound Descriptors for Impact Driving Broadband and M-weighted D1 SPL Peak

D2 RMS90

D3 SEL and Cumulative SEL

Appendix E Transmission Loss E1 Vibratory Driving E2 Impact Driving

1.0 Introduction

This report present the results of hydroacoustic monitoring which occurred in September and October, 2020 within the mouth of the Columbia River, near Chinook, Washington. The project was designed and overseen by the United States Army Corps of Engineers (USACE). Pacific Pile and Marine LLC (PPM) with Duwamish Services LLC performed the pile driving and related construction work. Robert Miner Dynamic Testing, Inc. (RMDT) carried out the hydroacoustic monitoring at the request of PPM.

The hydroacoustic monitoring was guided by a Hydroacoustic Monitoring Plan prepared by RMDT (August 31, 2020) and approved by the USACE. This plan addressed the requirements put forth in USAGE specifications. The Incidental Harassment Authorization (IHA) issued by the National Marine Fisheries Service (NMFS) did not directly address acoustic monitoring or sound source verification. In accordance with the Hydroacoustic Monitoring Plan, RMDT completed hydroacoustic monitoring as 24" OD piles were advanced with vibratory and impact hammers. Sound levels were monitored simultaneously at four distances from the piles during driving, and ambient underwater sound levels were monitored for three days after the pile driving ended.

1.1 Project Summary and Setting

The U.S. Army Corps of Engineers, Portland District (the Corps) conducted the Sand Island Test Pile Program near Chinook Washington on the north side of the Mouth of the Columbia River. Steel Test Piles will be driven along the alignments of existing Pile Dikes 4.01 and 6.37. All piles were 24" diameter steel pipe piles having a 0.50" wall thickness. Piles were supplied by the USACE. The Test Pile Program investigated driveability of these piles with four different tip treatments using a vibratory driver and an impact hammer. A total of 29 piles were called for on Plan Sheets C-101 & C-102, with thirteen at Dike 4.01 and sixteen at Dike 6.37. An open-end diesel impact hammer advanced most Test Piles; nine Test Piles were advanced with only a vibratory hammer.

Piles driven through the crest on an existing pile dike were advanced through surficial rip-rap rock fill and sometimes then into naturally occurring granular soils (primarily sand) below the rip-rap. Piles driven offset from the crest of the dike penetrated the granular river deposits. Final penetrations ranged from somewhat less than 18 ft, to slightly more than 42 ft.

1.2 Piles

All piles were 24" OD steel pipe pies with a wall thickness of 0.50-inches. Pile lengths ranged from 65 to 80 ft. Four different pile tip configurations were advanced. These tip configurations were open-ended without tip protection (“Pile Only”), open-ended with driving shoe (“Ring”), close-ended rock-breaking shoe (“Rock”), and close-ended conical tip (“Cone”). The location and tip treatment of each pile appears in Table 1.

-1- Robert Miner Dynamic Testing, Inc.

1.3 Hammers and Driving Operations

Two pile driving hammers were used on this project – one vibratory hammer and one impact hammer. Details for both hammers are summarized in Table 2. The APE 200 is a hydraulically powered vibratory hammer. The APE D50-52 is an open-end diesel hammer with variable fuel settings. For an open end diesel hammer such as the D50-52 the ram stroke height varies with the fuel setting and also with the soils resistance to driving; energy delivered to the pile is primarily a function of the ram stroke height.

2.0 Methods

The hydroacoustic monitoring followed guidance documents issued by the National Marine Fisheries Service (NMFS) in 2012 and the project Hydroacoustic Monitoring Plan dated August 31, 2020. Additionally, analysis of the measurements followed aspects of NMFS guidance issued in 2018.

2.1 Monitoring Locations

Hydrophones monitored underwater sound at various radial (horizontal) distances from the driven piles. These locations may be described as follows:

a) Near Source: One hydrophone was located at a nominal distance of 10 m from the piles. This location may be referred to as the Near Source (or Sound Source) depending on context; in some parts of this report data from this 10 m location may be identified as corresponding to “SLM1". (This designation of “SLM1" indicates the equipment used for the measurement, as discussed below.) The Near Source hydrophone was deployed from the crane barge operated by PPM for the pile driving. This barge provided a stable platform. When deploying this hydrophone we made water depth measurements and maintained the hydrophone at a depth that was approximately mid-depth at that location, with periodic adjustment for tide level. The hydrophone was held in position using a bottom anchor with the hydrophone buoyant above that anchor at mid-depth. When the current was heavy the hydrophone was also held in place with a line running from the barge to the anchor. Flow shields were used to reduce flow noise.

b) Intermediate Range 50 m: One hydrophone was placed at a nominal distance of 50m from the pile. This location may be referred to as the “50 m” location and may also be referred to as “SLM2" in this report. The 50 m hydrophone was deployed from crane barge operated by PPM for the pile driving. This barge provided a stable platform. When deploying this hydrophone we made water depth measurements and maintained the hydrophone at a depth that was approximately mid-depth at that location, with periodic adjustment for tide level. The hydrophone was held in position using a bottom anchor with the hydrophone buoyant above that anchor at mid-depth. When the current was heavy the hydrophone was also held in place with a line running from the barge to the anchor. Flow shields were used to reduce flow noise.

-2- Robert Miner Dynamic Testing, Inc.

c) Intermediate Range 100 m: Two hydrophone were placed near a nominal distance of 100 m from the pile. This location may be referred to as the “100 m” location.

These hydrophones were autonomous units deployed each day prior to pile driving and recovered each day after pile driving stopped. One hydrophone was placed approximately 1.5 m above the river bottom, and one was placed near the mid-depth expected for the period of recording. These autonomous recording hydrophones were buoyant above a primary anchor, with a recovery line consisting of sinking-line which extended 50 to 70 m laterally to a second anchor which secured a recovery buoy used for daily retrieval. Within this report the 100 m distance results are indicated by “LgH E” and “LgH F” for the mid-depth hydrophone and the hydrophone 1.5 m above the river bottom, respectively. The “LgH x” designation indicates use of a Loggerhead Snap unit, as discussed below.

d) Far Field: Two hydrophone were placed near a nominal distance of 500 m from the pile. This location may be referred to as the “500 m” location. These hydrophones were autonomous units deployed each day prior to pile driving and recovered each day after pile driving stopped. One hydrophone was placed approximately 1.5 m above the river bottom, and one was placed near the mid-depth expected for the period of recording. The deployment scheme was identical to that of the Intermediate 100 m units. Within this report the 500 m distance results are typically indicated by LgH C and LgH D for the mid-depth hydrophone and the hydrophone 1.5 m above the river bottom, respectively, where LgH denotes use of a Loggerhead Snap unit.

e) Background Sound Levels: An autonomous submersible hydrophone system was deployed approximately 2 m above the river bottom. This unit was deployed near the southern tip of West Sand Island approximately 600 m from the center of the Test Pile area at Dike 4.01. The deployment and retrieval scheme was identical to that of the Far Field described above, except that the unit was deployed for 72 hours prior to retrieval.

2.2 Monitoring Equipment

a) At the 10 m and 50 m locations underwater sounds were monitored using piezoelectric hydrophones manufactured by Reson. A Reson TC4040 hydrophone having a nominal receiving sensitivity of -206 dB re: 1 V/µPa measured sounds at the Near Source (10m) location. A Reson TC4033 unit was deployed at the 50 m location. The nominal receiving sensitivities of the TC 4033 is -203 dB. re:

1V/µPa. Signals from the TC4040 and TC 4033 piezoelectric units were charge conditioned using Bruel and Kjar Type 2635 Charge Amplifiers which, in turn, yielded signals with voltage proportional to sound pressure. Bruel and Kjar Type 2270 Class 1 sound level meters continuously digitized the signals with a 48 kHz sampling rate, and stored the signals (continuous wave-forms) in calibrated uncompressed wave files. The 2270 sound level meters also displayed real-time results for peak SPL and SEL computed for a 1 second period , thus providing some review of measurement quality.

-3- Robert Miner Dynamic Testing, Inc.

b) At the 100 m and 500 m locations we deployed units wherein the hydrophone directly connected to a digital acquisition system, all be submersible and autonomous. These units were Loggerhead Model SNAP units, manufactured by Loggerhead Instruments of Sarasota FL. The hydrophone elements were piezo-electric, with integral amplifier as manufactured by HTI in Long Beach MS. We used interchangeable hydrophone elements having receiving sensitivities of either -170 (2 each) , -180 (4 each) or -210 dB (2 each) re 1V/Pa.

Over the course of the project we deployed six different Snap units. The particular Snap/hydrophone combination depended on the anticipated sound levels at the respective locations. Each SNAP unit digitized the signal at 48 kHz and stored the data on micro-sd cards in uncompressed wave files.

c) Background sound levels were measured by a Loggerhead Snap unit with an amplified hydrophone having a receiving sensitivity of -170 dB re: 1V/Pa. This unit was deployed buoyant above an anchor as described for the Far Field measurements in Section 2.1 above. The SNAP unit recorded continuously for 72 hours, digitally storing wave files with sample rate of 48 kHz

d) High-pressure piston-phone calibrators, G.R.A.S. Model 42AC, provided calibrated sound sources to verify proper set up and function of the all hydrophones and recording equipment. During pile driving two piston-phones and couplers for all hydrophones were used on site each day. Calibration signals from these sound sources were recorded at the start and end of each day of recording so that proper sensitivity could be verified when reprocessing data. Similar piston-phone calibrations were made before and after making background measurement, except that calibration signals were made immediately before and after the 72 hour monitoring period rather than daily.

2.3 Measurement Sequence

Acoustic measurements of pile driving occurred on 15 days, with the first and last days being September 16 and October 9, respectively. Due to interruptions from weather standby, weekends and mechanical standby, the pile driving was not continuous.

During the field work we referred to days by date and by sequential day number. Table 8 provides a correlation of monitoring Day Number to calendar date.

The work in Pile Dike 6.37 began and ended September 16 and October 5; during the 15 days of pile driving in that period we monitored 20 installations by vibratory hammer, and 11 installations by impact hammer. Work in Pile Dike 4.01 encompassed 4 sequential days, ending on October 9. At Pile Dike 4.01 we monitored 12 vibratory installations and 10 impact installations. Two piles, 6-1C and 6-2C, were each monitored on two days because site conditions precluded completion of either pile in a single day; these piles appear twice in all of our monitoring data and tabulated results.

Most piles were initially installed with the vibratory hammer and then completed with an impact hammer. Eight pile were subject to only vibratory driving. Tables 1a and 1b

-4- Robert Miner Dynamic Testing, Inc.

rmsSPL T p t dt po

2 2log ( ) indicate the installation methods used for each pile.

Background measurements began near midday October 12 and ended 72 hours later on October 15. During this time there was no construction activity - the contractors floating equipment had been removed from the area.

2.4 Analyses

Continuous digital recordings of the time-series wave forms were reprocessed to review aspects of data quality and to compute sound descriptors appropriate for vibratory and impact driving. Consistent and correct amplitude during reprocessing were confirmed by reprocessing the daily field recordings of the calibrated piston-phone sound source.

The descriptors (metrics) used to characterize vibratory driving differ from those used to characterize impact driving, as discussed below.

2.4.1 Analyses for Vibratory Driving

For vibratory driving we present metrics which we now refer to as rms SPL and SEL.

These quantities are labeled differently in different documents. For clarity we note that rms SPL as referenced herein is computed as follows:

where the time duration, T, is the rms integration (“averaging”) period and p(t) is pressure at time, t. The reference pressure, p0, is 1µPa.

Specifications and the Monitoring Plan called for a 10 second RMS averaging period, that is, T =10 seconds. It is desirable that the rms SPL used to characterize hammer operation not include intervals when the hammer was not operating. Because hammer operation is subject to frequent starts and stops a simple series of consecutive 10 second periods will not generally exclude times when the hammer is off and also encompass all times when the hammer is operating. Thus, we computed the rms SPL for one second periods and then suitably summed the linear values with “hammer on” for 10 second periods. In this manner we were able to exclude from the rms SPL metric most one second periods when the hammer was not operating and include nearly all the time when the hammer was operating. In the presence of strong signals which clearly rise above ambient levels, such as at the Near Source, such data selection was unambiguous.

The metric referred to as the Sound Exposure Level, SEL, is the accumulated sound energy for the period, T100, expressed as an equivalent energy occurring for a period of 1 second, as indicated below wherein T0 is 1 second.

-5- Robert Miner Dynamic Testing, Inc.

SEL p t dt T p o o

10 10

2 2 log ( ) peakSPL p t po log max ( )

RMS

T p t dt p p t dt t t o

90 10 log

The SEL metric is expressed in dB re 1µPa2-second. As indicated in specific contexts the period, T100, may address the accumulated sound energy resulting from hammer operation for one pile, or one day, or some other stated period or event.

2.4.2 Analyses for Impact Driving

Sound descriptors for impact driving are computed for the pulse of sound energy associated with individual strikes of the hammer on the pile. The pulses from such strikes generally are contained within a time period of 80 to 150 ms and are typically separated from prior and subsequent pulses by a period of 1000 and 1400 ms.

The Peak SPL descriptor is the absolute value of the largest instantaneous pressure change in a single strike, expressed in dB re: 1µPa as follows:

For impact driving the root mean square (rms) SPL is computed for a time which encompasses 90% of the sound energy in a single strike. This sound descriptor may be referred to as 90% RMS SPL, SPLrms, RMS90, RMS90, etc. In this report we typically refer to this value as RMS90; it is computed as follows:

where the integration limits t05 and t95 are the times at which the energy in the pulse reaches 5 and 95 percent of the full pulse energy, respectively, and T90 is the length of time between t05 and t95 which therefore encompasses 90 percent of the pulse energy.

RMS90 is given in dB re 1µPa. Impact pulses typically have a concentration of energy in a short portion of the pulse, and thus reduction of the integration limits to encompass 90 percent of the energy increases the magnitude of the RMS90 descriptor because T90 is typically much less than 0.9 times the full pulse length.

The single strike sound exposure level, SELss,, is a measure of the sound energy in a pulse associated with a singe strike of the hammer on a pile. This sound descriptor is expressed in dB re 1µPa 2sec and computed as follows:

-6- Robert Miner Dynamic Testing, Inc.

SELss p t dt

T po o log cSEL SELss i

N

10 1010

10log where T is the full strike duration, and To is 1 second.

For impact driving the cumulative SEL, cSEL for a series of N strikes is expressed in dB re 1µPa 2sec and given by:

The sound descriptors mentioned above are computed for a specific frequency band, or after a specific auditory weighting function is applied to the signal. In this report the frequency band referred to as Broadband has been subject to high pass and low pass filters at 10 Hz, and 20 kHz, respectively. When auditory weighting is applied, the weighting follows the M-weighting for marine mammals as given in 2018 technical guidance from NMFS. These M-weighting filters distinguish between Cetaceans as Low Frequency (LF), Mid Frequency (MF), High Frequency (HF), and Phocid (PW) and Otariid (OW Pinnipeds in water. The M-weighting was applied during post-processing of the recorded broadband signal.

This report addresses underwater sound measurements. Thus, the reference pressure, p0, is 1µPa when expressing sound amplitude in decibels (dB). As such, the sound descriptors rmsSPL, peakSPL and RMS90 are expressed in dB re: 1µPa and SEL metrics are expressed in dB re: 1µPa 2sec. These dB references pressures (“re:

1µPa”) or reference exposures ( “re: 1µPa 2sec”) are consistent, and implied by the context, and are thus not herein repeated each time values are given for a quantity defined above expressed in decibels for either vibratory of impact pile driving.

2.4.3 Spectral Considerations

The measured sound levels may be considered to be the result of a combination of pressure variations which each have a different characteristic frequency. The absolute or relative sound energy within specific frequency limits may be investigated and quantified, and may be relevant to sound propagation as well as to the effects which the sound may have on marine mammals and other elements of interest. The computed Power Spectral Density (PSD) indicates the distribution of sound power over frequency with frequency divided into bins of 1 Hz width, and may be referred to as spectrogram or spectrograph. A similar presentation referred to as Octave Band analysis presents the variation of sound energy within bins having a width of an octave or some fixed fraction of an octave, such as one-third of an octave.

-7- Robert Miner Dynamic Testing, Inc.

TL B Log R

R

For this project field measurements of sound were recorded in the time-domain and spectral information was obtained from post-processing of the measured data. Per the specifications and the monitoring plan digital filters were applied to remove most sound energy at frequencies below and above 10 Hz, and 20 kHz, respectively. Thus, spectral results given herein are valid only within that range for pile driving. For background sound measurements the lower frequency cutoff was 15 Hz.

2.4.4 Transmission Loss

Transmission Loss is the decrease in acoustic intensity as an acoustic pressure wave propagates away from a source. The distance-related reduction of sound levels may result from a combination of geometric spreading, energy loss to the medium through which the sound propagates, and complex interaction with non-uniformities in the medium or transmitting or reflecting the propagating sound. Per the practical spreading loss model the NMFS considers that Transmission Loss, TL may be calculated as follows, where the units of TL are dB, B is a transmission loss coefficient, R1 is the distance between the point of model interest and the pile, and R2 is the distance between the driven pile and the point of the assumed Source level.

The Transmission Loss equation may be arranged to evaluate the loss coefficient, B, given simultaneous measurements at two or more various radii (distances); this is typically done using regression methods and data for three or more distances. The equation may also be arranged to compute the distance, R1, to a particular injury or disturbance threshold level given a measured or assumed source level at R2 and an assumed Transmission coefficient, B.

3.0 Results

The primary purposes of this document are to report sound source levels during vibratory and impact driving at this site and background sound levels without construction activity. Additional purposes include characterizing Transmission Loss and reporting empirically supported distances to Level A and Level B injury/harassment threshold sound levels.

3.1 Characteristic Waveforms and Spectra

Figures 1 and 2 depict characteristic waveforms of measured pressure versus time for vibratory and impact pile driving on 24" OD piles driven through rock rip-rap on the crest of Dike 6.37, Location 6-1C. Figure 1 depicts a short period of time beginning with the APE 200 vibratory hammer operating such that little energy was applied and ending with normal operation. The latter portion of Figure 1 indicates a cycle time of

-8- Robert Miner Dynamic Testing, Inc.

approximately 40 ms, corresponding to an apparent hammer oscillation at approximately 26 Hz. Figure 2 depicts a single strike event with the APE D50-52 impact hammer; the strike event has a duration of approximately 65 ms in which the measured pressure varies quite rapidly before returning to relative quiescence pending the next hammer strike. The time between individual strikes was typically 1.2 to 1.5 seconds such that the strike rate was usually between 40 and 50 blows per minute with a characteristic rate of 45 blows per minute.

Figure 3 depicts the computed spectra for the vibratory driving waveform depicted in Figure 1. Figure 4 depicts the computed spectra for the single strike impact driving waveform shown in Figure 2. From these figures it may be observed that the majority of sound energy was present at frequencies below 5 kHz.

The sound descriptors presented in Sections 2.5.1 and 2.5.2 are computed for a specific frequency band, or after a specific auditory weighting function is applied to the measured signal. In this report the frequency band referred to as Broadband has been subject to high pass and low pass filters at 10 and 20 kHz, respectively. (Except for Background measurements for which the high pass filter was set at 15 Hz.) When an auditory weighting was applied, such weighting followed the M-weighting for marine mammals as given in 2018 technical guidance from NMFS. These M-weighting filters apply to cetaceans as Low Frequency (LF), Mid Frequency (MF), High Frequency (HF), and to Phocid Pinnipeds (PW) and Otariid Pinnipeds (OW). Figures 5 and 6 present the computed PSD values after application of the various M-weighting filters to the waveforms measured at 100 m and shown in Figures 1 and 2 for vibratory and impact driving. Figure 7 depicts the effect which the filters have on computed the PSD for a single impact strike measured at 10 m. Sound descriptors values computed for the Near Source after application of the M-weighted filters to measurements of vibratory and impact driving are in Appendix C and D, respectively.

At the 100 and 500 m locations measurements were made at two depths as called for in project documents. One Loggerhead unit was at mid depth, and one was approximately 1.5 m above the river bottom. Characteristic results for a single impact hammer strike as measured by both hydrophones are presented in Figures 8 and 9 for the 100 and the 500 m locations, respectively. The measured results are quite similar at the two hydrophone depths, particularly with respect to the frequency content as indicated in the PSD. During data reduction we noted that averaged or median results for these two depth were similar, but the hydrophone located 1.5 m above the bottom occassionally had more variation. This variation was more apparent at the 500 m distance where sound levels were lower. In consequence of this noted variation, and the general consistency of median and average values we used the mid-depth data for characterizing levels applicable to Transmission Loss analysis..

Further characteristic broadband waveforms, associated PSD plots, and additional signal characterizations appear in Appendix A1 and A2 for a sampling of vibratory and impact events.

-9- Robert Miner Dynamic Testing, Inc.

3.2 Sound Levels for Vibratory Driving

Sound descriptors which characterize vibratory driving on this project, are the rmsSPL for a 10 second period, and the cumulative SEL. The Source Sound levels discussed in this Section 3.2 are those for the nominal 10 m location and are herein referred to as the Near Source or Source levels. (Results for greater distances are in Appendix C and are discussed in later sections.) The Near Source results are presented in several contexts as discussed below.

3.2.1 Broadband Near Source Levels, Vibratory Driving

Table 3 summarizes the computed broadband (10 Hz to 20kHz) results for each test session 10m from each monitored pile. A summary of rms SPL results is given for each test together with the computed SEL for each session and the total time the vibratory operated to drive each pile. The overall median and mean values at Pile Dike 6.37 for rms SPL were 167.8 and 1167.2 dB, respectively. The overall median and mean values at Pile Dike 4.01 for rms SPL both 170.0 dB, respectively. These overall values are the averages giving equal weight to each individual pile. Driving times represent time of hammer operation rounded to the nearest 10 second interval; times for various interruptions are excluded.

The broadband cumulative daily SEL values ranged from 171 to 204 dB at Pile Dike 6.37and ranged from 190 to 203 at Pile Dike 4.01. This relatively wide range of results reflects wide variation in the nature of the activity. These cumulative daily SEL values were computed using only those time intervals during which the hammer was operating;

the Daily Total Drive Time is the span from the start of driving to the end of driving, exclusive of the time the hammer was not in operation.

For each day that we monitored vibratory driving, Appendix B1 contains a “Session Log” plot of the rms SPL and accumulating SEL versus time for vibratory driving. Transitions between piles are marked with labels on the plots.

3.2.2 M-Weighted Near Source Levels, Vibratory Driving

For each day of vibratory monitoring Appendix C contains as summary of the rms SPL values computed after application of five different standardized marine mammal auditory weighting functions. The M-weighted results are markedly lower than the broadband results.

3.3 Near Source (10 m) Sound Levels for Impact Driving

Sound descriptors for measurements of impact driving on this project are the peakSPL, RMS90, single strike SEL, SELss and the Cumulative SEL, cSEL. Near Source measurement results are presented in this Section 3.3 ; results for greater distances are in Appendix D and are discussed in later sections.

3.3.1 Broadband Near Source Levels, Impact Driving

Tables 4, 5 and 6 summarize the computed broadband results for the metrics applicable to single strike analyses, namely, peakSPL, RMS90, SELss. Table 6 which summarizes

-10- Robert Miner Dynamic Testing, Inc.

the single strike SEL values also gives the daily cumulative SEL, cSEL, resulting from impact driving. Appendix C contains more detailed summaries of the computed metrics.

Measurements of the peakSPL metric at Pile Dike 6.37 yielded an average and median value of approximately 202 dB and 201 dB respectively. Mean RMS90 values ranged from 183 to 197 dB, with an overall average of 188 dB. Single Strike mean SEL values ranged from 169 and 181 dB.

Measurements of the peakSPL metric at Pile Dike 4.01 yielded an average and median value of approximately 204 dB and 204 dB respectively. Mean RMS90 values ranged from 186 to 194 dB, with an overall average of 191 dB. Single Strike mean SEL values ranged from 173 and 180 dB.

Cumulative (daily) Sound Exposure level, cSEL, varied widely depending on the extent of impact driving during a particular day. The results ranged from 182 to 205 dB, respectively when considering both Pile Dike locations.

The single strike metrics (peak SPL, RMS90 and SEL) computed for each day we monitored impact driving are plotted for sequential strikes on sheets in Appendix B2.

Also included on each “Session Log” plot of impact driving are the values for cumulative SEL, cSEL, as driving progressed, with the horizontal axis as strike number.

3.3.2 M-Weighted Near Source Levels, Impact Driving

Marine mammal auditory weighting functions were applied during post-processing of impact driving and Appendix D contains the results for M-weighted Near Source sound levels during impact pile driving.

-11- Robert Miner Dynamic Testing, Inc.

3.4 Transmission Loss

Documents for this project call for calculation of the observed Transmission Loss. For this purpose we assembled measured broadband results measured at various distances for the same events. The data available for this consisted of measurements at nominal distances of 10, 50 , 100 and 500 m. Using these data the Transmission Loss coefficient, B, is obtained by plotting measured sound levels at various distances and computing the slope of a regression line for those points on a semi-log plot where distance is on a log10 scale. The absolute value of the slope of the regression line is the apparent Transmission Loss coefficient, B, and it indicates the apparent decibel reduction per tenfold increase in radial distance from the source pile.

For practical reasons the actual distances from the piles sometimes differed from the nominal distances. The practical reasons which caused some variation in distances include drift above a stable anchor due to tidal current change, deployment decisions made to minimize movement related to current and wave, risk of interference/overlap with other public or construction activity and effort to avoid shallow water or highly variable bathymetry between the pile and the hydrophone. At the Near Source location, actual distances were typically 10 or 11 m, but distances were 7 m, and 9 m on 2 and 6 occasions, respectively. The nominal 50 m location was typically 60 m, was twice 65 m, and 4 times was 45 m. The Nominal 100 m location was typically between 100 and 130 m, and was 140 m on two occasions. The nominal 500 m location was most often 500 m, was 530 to 550 m on ten occasions and was 560 or 580 m on six occasions. These varied actual distances were used in Transmission Loss calculations, and are apparent in the Transmission Loss plots.

Appendix E contains the Transmission Loss plots for each day separated based on hammer type, with vibratory or impact driving in part E1 and E2, respectively. Each plot is a semi-log plot is shown together with the best fit regression line and the computed slope, B. For the regression calculation of vibratory driving we used the rmsSPL versus radial distance. For regression of the impact data we used the RMS90 descriptor, but included the peakSPL and SELss data on the plots for reference. Twenty five data sets were assembled representing daily results for the two different methods of driving, vibratory and impact.

Twenty-three of these plots incorporate measurements at four different radial distances.

Plots for vibratory driving on Day 4 and impact driving on Day 5 incorporated data for three locations because on those days sound levels exceeded the maximum levels we anticipated for either the 100 m or 500 m locations on those days, and the recordings are flawed due to over-range. For the nominal 100 and 500 m locations our field data includes measurements at mid-depth and 1.5 m above the river bottom. The Transmission Loss calculations use only the mid-depth measurements such that the 100 and 500 m locations do not take disproportionate weight in the regression, and also so that all data was for the same mid-depth position. Table 8 summarizes the computed Transmission Loss Coefficient values.

-12- Robert Miner Dynamic Testing, Inc.

3.5 Background Sound Levels

Figure 7 presents the measured background sound measurements of a three day period in mid October, 2020. The results are presented as rms SPL for a 30 second integration period and are computed from measurements having a frequency range of 15 Hz to 20 kHz. The diurnal tidal fluctuation is plotted in Figure 7 in terms of tidal water height at nearby Cape Disappointment. Figure 7 also includes the observed wave height nearby.

Changes in the wave height do not appear have significant influence on background sound in this case. However, there does appear to be a strong diurnal variation in background sound. Figure 8 presents the Cumulative Distribution Curve for the broadband (15 Hz to 20 kHz) measurements and also the M-weighted measurements.

The median value for the broadband 30 second RMS level is 114.3 dB re: 1µPa.

4.0 Discussion and Summary

4.1 Near Source Sound Levels

4.1.1 Vibratory Pile Driving

The results presented herein address underwater sound levels measured over approximately 703 minutes of vibratory hammer operation advancing 24" diameter steel piles in water depths ranging from 1 to 7 ft. The data covers vibratory driving on 32 different piles advanced on 32 occasions spread over 15 days of monitoring. For the broadband measurements at a distance of 10 m from the source pile Table 3 presents statistics regarding the rmsSPL computed for a 10 second averaging time for each pile;

the overall project median, mean and standard deviation values are 170.7, 168.6 and

8.5 dB re 1µPa, respectively. Daily cumulative SEL values for broadband ranged from 171 to 203 dB re 1µPa 2s.

Appendix C contains results for the 10 m distance after application of the marine mammal auditory weighting functions (M-weighting filters). The LF, MF, HF, filters reduced the Near Source rmsSPL by approximately 2, 15 and 12 dB, and the PW and OW filters reduced rmsSPL by approximately 5 dB.

4.1.2 Impact Pile Driving

Monitoring during impact driving addressed 10 days of driving during which 21 24" OD piles were advanced with an open-end diesel impact hammer. The overall median and mean RMS90 were both 190 dB per strike, with a total of 2779 hammer strikes. These results were summarized with further detail in Table 5a and 5b. The daily cumulative SEL, cSEL, ranged from 182 to 205 dB re 1µPa 2s.

Appendix D contains results for a Near Source (10 m) after application of M-weighted filters. The MF, HF, filters reduced the Near Source rmsSPL by approximately 30 and 45 dB, and the PW and OW filters reduced rmsSPL by approximately 33 dB.

-13- Robert Miner Dynamic Testing, Inc.

4.2 Transmission Loss

We computed a Transmission Loss coefficient, B, for data obtained during 15 days of vibratory driving and 10 days of impact driving data. Significant variability was observed. However, within this data we did not observe patterns which would suggest that the pile type (open or closed end) effected Transmission loss. For vibratory and impact driving the average Transmission Loss coefficients were 16.7 and 18.7 dB, respectively, and the overall average for daily data was 17.5 dB.

The average Transmission Loss coefficient values computed for this project are higher than the commonly assumed default value of 15 dB. However, measurements completed by RMDT approximately 2 miles south of the site, at the South Jetty Material Offloading Facility yielded a Value of 16.1 dB during vibratory driving in June of 2020.

While the reason that Transmission Loss computed for this project are slightly higher than the commonly assumed value of 15 are not known, two potential causes may be postulated. One potential cause pertains to the fact that most of the pile driving occurred with pile tips in rock rip-rap at relatively shallow depth on or near the crest of the rock berm. Such driving may be considered unusual, and is it reasonable to conjecture that the locus of the source sound energy is unusual and that relatively less energy would reach the underling granular soil to propagate laterally as is the normal case. Such a condition might be expected to cause relatively higher sound levels at the closer measurements and relatively lower sound levels at distant locations.

The second potential cause for higher project Transmission Loss pertains to the relatively modest water depth; from principle and experience it is known that lower frequency components in the source sound are attenuated by shallow water. This behavior is apparent when comparing the PSD results for various distances. Figure 12 depicts the relative frequency content of signals measured at distances of 10, 50 , 100 and 500 m. In Figure 12 the attenuation with distance is greatest for frequencies below approximately 50 Hz.

4.3 Computed Distances to Sound Thresholds

The sound source characterizations and Transmission Loss coefficients computed for this project may be used to calculate the distances to Level A Injury and Level B disturbance for the various marine mammal functional hearing groups.

4.3.1 Vibratory Pile Driving

For non-impulsive sound which results from vibratory pile driving the Level A isopleths for Auditory Injury Thresholds (PTS) are given in terms of cumulative SEL. We computed the distance to these thresholds using the M-weighted cSEL values taken from the monitoring day having the largest computed cSEL. This day was Day 12 during which the vibratory hammer operated for a total of 2 hours to drive 3 piles. For each marine mammal hearing group (LF, HF, MF, PW and OW) Table 8 lists the Auditory Injury Thresholds and the computed distances to these thresholds. For High Frequency Cetaceans the computed distance was 191 m. For all other hearing groups the distance was less 28 m.

-14- Robert Miner Dynamic Testing, Inc.

We computed the distances to the Level B isopleth using as the Sound Source the median RMS value for Dike 4.01. (The median RMS value for Dike 6.37 was lower than the 170 dB value for Dike 4.01.) Table 9 lists the Behavioral Disturbance Threshold and the computed distance of 9870 m to this Disturbance Threshold during vibratory driving.

4.3.2 Impact Pile Driving

For impulsive sound, such as the sound from impact pile driving, the Underwater Noise Thresholds for Auditory Injury (PTS) are given in terms of M-weighted cumulative SEL values for the five functional hearing groups, and a broadband peak SPL for each hearing group. The M-weighted cumulative SEL values we used to compute the threshold distances were taken from the single monitoring day having the largest cSEL value for impact driving. The peak SPL value was taken to be largest daily median value for the project. Results are given in Tables 8 and 9.

For impact driving the distance to the Injury Threshold was computed to be less than 25 m. This very modest distance to the Injury Threshold primarily reflects that fact that impact pile driving was of relatively short duration on this project; the field day with the highest cumulative SEL had only 413 hammer strikes. The computed distance to the disturbance Threshold was 708 m for impact driving.

Table 9 gives the calculated distances to Injury Thresholds for fish subject to sounds from impact pile driving. Using the measured field data these distances range from 23 to 32 m.

4.3.3 Extension of Results

The calculation results given in Tables 8, 9 and 10 for marine mammals and fish reflect the measurements taken during the Test Pile Program. If conditions differ during production driving, then distances to thresholds will differ. In our opinion, the most relevant potential difference pertains to the impact driving Level A threshold; if the number of pile strikes increases, the distance to the threshold will increase. However, based on the results of the Test Pile Program, it may be expected that distances to Injury Thresholds will remain quite modest during impact driving, even if the number of strike increases quite markedly.

The distances to Disturbance Thresholds are based on the characteristic sound source in terms of RMS90 for impact driving, and in terms of RMS for 10 second integration time during vibratory driving. Such sound source values do not generally increase when the quantity of strikes or duration of driving increases.

-15- Robert Miner Dynamic Testing, Inc.

4.4 Background Sound Measurements

Figure 7, 8 and 9 present results from the background measurements near the tip of West Sand Island and near the site of Dike 4.01. Considering the 72 hour period, 114.3 dB re: 1µPa was the median value of the RMS calculation for a 30 second period using the broadband (15 Hz to 20 kHz) data.

These data for West Sand Island in mid-October are relatively consistent with measurements made by RMDT approximately 2 miles from this site during May of 2020

- on that occasion the measurements showed a strong diurnal variation, with a median value of 118.2 dB for a 30 second RMS level of a period of 72 hours. Those prior measurements were made with a Reson TC4032 hydrophone which is in practice more sensitive that the HTI unit used at Sand Island. Our review of both measurements, and also the character of the cumulative distribution function for the present data, suggest that background levels at West Sand island may have occasionally dropped below the noise floor of the equipment we deployed at West Sand Island. If this is the case, then the present median value of approximately 114 dB may be slightly higher than the actual median value in situ. The autonomous and fully submersible Loggerhead Snap unit deployed at West Sand Island provided satisfactory measurements despite the very harsh conditions with wave heights exceeding 12 ft. The system based on a Reson TC 4032 could not have been deployed with satisfactory results in these harsh conditions.

Per the PSD and one-third Octave band results plotted in Figure 9, frequencies above approximately 500 to 1000 Hz provided a very small portion of the sound energy. This result is consistent with our results from May 2020 on the south shore of the MCR.

-16- Robert Miner Dynamic Testing, Inc.

Table 1a: Pile Locations &Test Details, Dike 6.37, September 16- October 2, 2020

Pile Identification Location:

Dike and Station

Position Relative Existing

Rock Berm

Installation Method

Pile Length ft

Approx.

Final Depth Below

Mudline ft

6-1C Rock 6.37 Sta 33+00 On Crest Vib/Impact 65 21

6-1C Ring 6.37 Sta 33+00 On Crest Vib/Impact 65 36

6-1F Pile Only V 6.37 Sta 33+00 Offset Vib/Impact 80 8

6-1F Cone V 6.37 Sta 33+00 Offset Vib 80 36

6-2C Cone 6.37 Sta 34+80 On Crest Vib/Impact 70 36

6-2C Rock 6.37 Sta 34+80 On Crest Vib/Impact 65 36

6-2F Ring 6.37 Sta 36+85 Offset Vib/Impact 80 43

6-2F Cone 6.37 Sta 36+85 Offset Vib/Impact 70 36

6-3C Rock 6.37 Sta 36+85 On Crest Vib/Impact 65 36

6-3C Ring 6.37 Sta 36+85 On Crest Vib/Impact 80 40

6-3F Pile Only 6.37 Sta 38+40 Offset Vib 80 36

6-3F Ring V 6.37 Sta 38+40 Offset Vib 80 36

6-3F Cone V 6.37 Sta 38+40 Offset Vib 80 32

6-4C Pile Only V 6.37 Sta 38+40 On Crest Vib 80 16

6-4C Ring 6.37 St 38+40 On Crest Vib/Impact 65 36

6-4C Cone 6.37 St 38+40 On Crest Vib/Impact 60 21

-17- Robert Miner Dynamic Testing, Inc.

Table 1b. Pile Locations and Test Details, Dyke 4.01, October 6-9, 2020

Pile Identification Location:

Dike and Station

Position Relative Existing

Rock Berm

Installation Method

Pile Length ft

Approx.

Final Depth Below

Mudline ft

4-1C Cone V 4.01 Sta 10+20 On Crest Vib 70 9.5

4-1C Cone 4.01 Sta 10+20 On Crest Vib/Impact 70 39

4-1F Pile Only V 4.01 Sta 10+20 Offset Vib 80 38

4-1F Ring 4.01 Sta 10+20 Offset Vib/Impact 80 38

4-1F Cone 4.01 Sta 10+20 Offset Vib/Impact 70 41

4-2C Ring 4.01 Sta 11+60 On Crest Vib/Impact 65 21.5

4-2C Cone 4.01 Sta 11+60 On Crest Vib/Impact 70 24

4-2F Cone V 4.01 Sta 12+60 Offset Vib 70 24

4-2F Cone 4.01 Sta 12+60 Offset Vib/Impact 70 39

4-3C Ring 4.01 Sta 12+60 On Crest Vib/Impact 65 38

4-3C Cone 4.01 Sta 12+60 On Crest Vib/Impact 70 42

4-4C Cone 4.01 Sta 13+70 On Crest Vib/Impact 70 41

4-4C Rock 4.01 Sta 13+70 On Crest Vib/Impact 65 42

Table 2. Summary of Driving System Details, Vibratory and Impact Hammers

Vibratory Hammer

Eccentric Moment

Frequency Range

Driving Force

Max Amplitude Free

APE 200 6,600 inch-lbs ~21 to 28 Hz 510 kips 1.17 inches

Impact Hammer

Max. Rated Energy kip-ft

Ram Weight kips

Maximum Ram

Stroke ft

Helmet Weight kips

Cushion Stiffness kip/inch

APE D50-52 124 11 11.3 4.69 39,981

-18- Robert Miner Dynamic Testing, Inc.

Table 3a. Pile Dike 6.37 Vibratory Driving, Statistics for Broadband rms SPL, 10 second integration period, and Cumulative SEL for a 10 m radial distance from the pile.

Pile (a)

Date

Day Depth, m (b) Median dB

Mean dB

Standard Deviation dB

Min.

dB

Max.

dB

Duration mm:ss

Daily Total Drive Time (hh:mm:ss) and Cumulative SEL

Water Pile Tip

6-3F PO V Sept 16 Day 1 5 36 160.6 160.6 1.3 159 162.3 0:01:50 171.2 dB

6-3F Cone V Sept 17 Day 2 8 11 178.4 178.3 3.1 171.7 183.7 0:17:50 0:42:00

199.7 dB 6-3F Ring V Sept 17 Day 2 8 36 163 156.1 15.8 124.3 173.4 0:09:40

6-4C PO V Sept 17 Day 2 1 16 162.8 162.3 3.1 153.4 167.6 0:14:30

6-1C Rock V Sept 18 Day 3 2.0 Na 173.4 172.9 1.8 166.6 176.8 0:37:50 2:00:40

204.1 dB 6-1F Cone V Sept 18 Day 3 6 36 170.9 173.6 5.6 126.4 185.2 1:09:00

6-1F PO V Sept 18 Day 3 1 8 159.4 159.4 4.7 146.1 168.8 0:13:50

6-1C Rock Sept 19 Day 4 2.0 Na 172.1 170.3 5.9 154.3 177.3 0:15:00 192.1 dB

6-1C Rock Sept 21 Day 5 2.0 Na 171.7 171.8 1.1 165.7 174.3 0:15:30 191.7dB

6-2F Cone Sept 22 Day 6 2.9 15 175 175.2 2.3 170.4 180.7 0:14:10 0:57:50

197.16-3C Spud Sept 22 Day 6 6.7 15 168.4 168.3 1.5 163.9 171.3 0:43:40

6-3C Ring Sept 27 Day 7 6.